On June 10, during its Worldwide Developers Conference (WWDC), Apple unveiled “Apple Intelligence,” a suite of AI-powered tools embedded in the next-generation platforms: iOS 18 for iPhone, iPadOS 18 for iPad, and macOS Sequoia for Macs. However, the subsequent headlines focused on Apple’s partnership with OpenAI for integrating ChatGPT into its intelligence suite. This development has sparked curiosity among technology enthusiasts about how the Apple-OpenAI partnership will function, given Apple’s emphasis on privacy in its intelligence systems. Additionally, there is speculation about potential collaborations with other technology companies like Google to enhance AI capabilities further.

Apple Intelligence and Privacy

Apple describes its Apple Intelligence as a “personal intelligence system” that integrates generative AI capabilities into iPhones, iPads, and Macs, with a primary focus on user privacy. Apple Intelligence utilises the on-device processing power of M-series Apple Silicon for Macs and iPads and the A17 Pro chip for iPhones. However, requests for advanced features requiring more processing power are handled in the cloud. For cloud processing, Apple said it has developed Private Cloud Compute (PCC), which ensures that user data is accessible only to the user and not even to Apple. Apple states in its security blog that once a response to a user’s request is sent, “personal data leaves no trace in the PCC system.”

Apple’s partnership with OpenAI

In an interview with Fast Company, Craig Federighi, Apple Senior Vice President of Software Engineering, explained how OpenAI’s ChatGPT has been integrated into the Apple ecosystem while maintaining user privacy. Federighi clarified that OpenAI’s ChatGPT does not power Apple Intelligence; instead, both systems independently provide AI capabilities on the same device. Apple Intelligence is driven exclusively by Apple’s LLMs and AI models, with no data sent to OpenAI without user permission. For instance, to receive ChatGPT-powered responses through Apple’s voice assistant Siri, users must permit Siri to send a request to ChatGPT.

OpenAI, in a blog post detailing its partnership with Apple, confirmed that requests to ChatGPT from Siri and other Apple AI tools are protected. OpenAI stated that such requests are not stored, and users’ IP addresses are obscured.

A potential deal with Google

While ChatGPT integration with Apple’s operating system will be available to users with iOS 18, Federighi indicated that Apple might collaborate with other large language model (LLM) providers in the future. According to 9To5Mac, he said, “We think ultimately people are going to have a preference perhaps for certain models that they want to use, maybe one that’s great for creative writing or one that they prefer for coding. And so we want to enable users ultimately to bring a model of their choice.” He specifically mentioned that Apple could integrate models like Google Gemini in the future.

Although Apple has not yet formalised a deal with Google for Gemini AI integration, it is already utilising Google’s technology to power its AI models. A report by Reuters indicated that according to a technical document released by Apple, the company’s engineers used Google’s cloud-based tensor processing units (TPUs) to train Apple’s foundational AI model. While Apple has not detailed the extent of its reliance on Google’s chips and software, it is likely that both companies have been collaborating in the AI space for several months.

First Published: Jun 12 2024 | 4:50 PM IST